Abstract
Conhecer o perfil das mulheres brasileiras que atuam em Processamento de Linguagem Natural (PLN) é um importante passo para o desenvolvimento de políticas e programas que visem aumentar a inclusão e a diversidade nessa área. Este é o primeiro trabalho realizado no Brasil com este fim. A partir de dados coletados via consulta pública, Lattes e Linkedin, notou-se que o perfil é de uma formação em computação ou linguística, atuando em empresas ou universidades, mas com pouca diversidade étnica e aparente dificuldade em conciliar vida profissional e maternidade. Analisando mais especificamente o grupo “Brasileiras em PLN” constatou-se uma expressiva capacidade de publicação e orientação, mas ainda uma baixa colaboração entre nossas integrantes.

Abstract
The estimation of the frequency of sinusoids has been the object of intense research for more than 40 years. Its importance in classical fields such as telecommunications, instrumentation, and medicine has been extended to numerous specific signal processing applications involving, for example, speech, audio, and music processing. In many cases, these applications run in real-time and, thus, require accurate, fast, and low-complexity algorithms. Taking the normalized Cramer-Rao lower bound as a reference, this paper evaluates the relative performance of nine non-iterative discrete Fourier transform-based individual sinusoid frequency estimators when the target sinusoid is affected by full-bandwidth quasi-harmonic interference, in addition to stationary noise. Three levels of the quasi-harmonic interference severity are considered: no harmonic interference, mild harmonic interference, and strong harmonic interference. Moreover, the harmonic interference is amplitude-modulated and frequency-modulated reflecting real-world conditions, e.g., in singing and musical chords. Results are presented for when the Signal-to-Noise Ratio varies between -10 dB and 70 dB, and they reveal that the relative performance of different frequency estimators depends on the SNR and on the selectivity and leakage of the window that is used, but also changes drastically as a function of the severity of the quasi-harmonic interference. In particular, when this interference is strong, the performance curves of the majority of the tested frequency estimators collapse to a few trends around and above 0.4% of the DFT bin width.

Abstract
Cricket sounds are usually regarded as pleasant and, thus, can be used as suitable test signals in psychoacoustic experiments assessing the human listening acuity to specific temporal and spectral features. In addition, the simple structure of cricket sounds makes them prone to reverse engineering such that they can be analyzed and re-synthesized with desired alterations in their defining parameters. This paper describes cricket sounds from a parametric point of view, characterizes their main temporal and spectral features, namely jitter, shimmer and frequency sweeps, and explains a re-synthesis process generating modified natural cricket sounds. These are subsequently used in listening tests helping to shed light on the sound identification and discrimination capabilities of humans that are important, for example, in voice recognition. © 2023 IEEE.

Abstract
Censored data are frequently found in diverse fields including environmental monitoring, medicine, economics and social sciences. Censoring occurs when observations are available only for a restricted range, e.g., due to a detection limit. Ignoring censoring produces biased estimates and unreliable statistical inference. The aim of this work is to contribute to the modelling of time series of counts under censoring using convolution closed infinitely divisible (CCID) models. The emphasis is on estimation and inference problems, using Bayesian approaches with Approximate Bayesian Computation (ABC) and Gibbs sampler with Data Augmentation (GDA) algorithms.

Abstract
Greenland ice core records display abrupt transitions, designated as Dansgaard-Oeschger (DO) events, characterised by episodes of rapid warming (typically decades) followed by a slower cooling. The identification of abrupt transitions is hindered by the typical low resolution and small size of paleoclimate records, and their significant temporal variability. Furthermore, the amplitude and duration of the DO events varies substantially along the last glacial period, which further hinders the objective identification of abrupt transitions from ice core records Automatic, purely data-driven methods, have the potential to foster the identification of abrupt transitions in palaeoclimate time series in an objective way, complementing the traditional identification of transitions by visual inspection of the time series.In this study we apply an algorithmic time series method, the Matrix Profile approach, to the analysis of the NGRIP Greenland ice core record, focusing on:- the ability of the method to retrieve in an automatic way abrupt transitions, by comparing the anomalies identified by the matrix profile method with the expert-based identification of DO events;- the characterisation of DO events, by classifying DO events in terms of shape and identifying events with similar warming/cooling temporal patternThe results for the NGRIP time series show that the matrix profile approach struggles to retrieve all the abrupt transitions that are identified by experts as DO events, the main limitation arising from the diversity in length of DO events and the method’s dependence on fixed-size sub-sequences within the time series. However, the matrix profile method is able to characterise the similarity of shape patterns between DO events in an objective and consistent way.

Abstract
Recently, learning-based image compression has attracted a lot of attention, leading to the development of a new JPEG AI standard based on neural networks. Typically, this type of coding solution has much lower encoding complexity compared to conventional coding standards such as HEVC and VVC (Intra mode) but has much higher decoding complexity. Therefore, to promote the wide adoption of learning-based image compression, especially to resource-constrained (such as mobile) devices, it is important to achieve lower decoding complexity even if at the cost of some coding efficiency. This paper proposes a complexity scalable decoder that can control the decoding complexity by proposing a novel procedure to learn the filters of the convolutional layers at the decoder by varying the number of channels at each layer, effectively having simple to more complex decoding networks. A regularization loss is employed with pruning after training to obtain a set of scalable layers, which may use more or fewer channels depending on the complexity budget. Experimental results show that complexity can be significantly reduced while still allowing a competitive rate-distortion performance.